You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm encountering a critical limitation when trying to re-add a previously dropped property in NebulaGraph (v3.8.0). Here's the scenario: 1.Current Behavior:
CREATE TAG person(name string, age int);
INSERT VERTEX person VALUES "1":("Tom", 23);
ALTER TAG person DROP (age); -- Schema version incremented
ALTER TAG person ADD (age string); -- Fails! Cannot re-add dropped property
Once a property (e.g., age) is dropped, re-adding it with the same name (even with the same type) is permanently blocked.
Workarounds like using new property names (age_v2) require application-layer changes, which is unsustainable for our production systems. 2.Pain Points:
Zero-Downtime Schema Evolution: In agile development, we may need to temporarily remove and later restore properties without breaking existing queries.
Data Recovery: Accidentally dropped properties cannot be gracefully recovered.
Business Logic Dependency: Some legacy applications hardcode property names (e.g., age), making workarounds costly. 3.Questions:
Is this limitation fundamental to LSM-Tree, or could it be relaxed with schema management improvements?
Would the team consider supporting this feature in future releases?
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
-
I'm encountering a critical limitation when trying to re-add a previously dropped property in NebulaGraph (v3.8.0). Here's the scenario:
1.Current Behavior:
Once a property (e.g., age) is dropped, re-adding it with the same name (even with the same type) is permanently blocked.
Workarounds like using new property names (age_v2) require application-layer changes, which is unsustainable for our production systems.
2.Pain Points:
Zero-Downtime Schema Evolution: In agile development, we may need to temporarily remove and later restore properties without breaking existing queries.
Data Recovery: Accidentally dropped properties cannot be gracefully recovered.
Business Logic Dependency: Some legacy applications hardcode property names (e.g., age), making workarounds costly.
3.Questions:
Is this limitation fundamental to LSM-Tree, or could it be relaxed with schema management improvements?
Would the team consider supporting this feature in future releases?
References:
#5130
Appreciate your insights!
Beta Was this translation helpful? Give feedback.
All reactions